Forums » General

Spin Ratios

«12
Nov 20, 2013 Kierky link
Okay, so what you mean by length is just a pretty number given to the client to make it look like we have a stat. Lol, so instead the client is working with a different number entirely. Perhaps make the stats actually correspond to the ship?

edit: yes TRS.
Nov 20, 2013 TheRedSpy link
Yeah its funny that all this time we actually thought that the length stat had gameplay impact i'm surprised nobody has asked the question as to whether all the stats actually mean something.

No wonder you were all n00bs when I came along!

Also i believe kierky meant to change the stat to correspond to the actual ship, not the ship to the stat.

Anyway the important thing with the spin ratio is not that it's accurate. Nobody is going to care to calculate how many seconds it would take to turn 90 degrees in a skyprom or whatever, the point is for comparative purposes all we need is a formula that is uniformly applied to all the ships so we can see relatively how fast each turns relative to the other.

Here's a question, if you expand the mesh with this new dynamic addon feature, you know the one you use for the type-S and stuff, does that mean that hypothetically you could completely change the spin mechanics of a ship by say adding more length dynamically or sticking on a dingbat? If you do this to all variants I spose we're going to get a huge deviation of manuverability amongst variants.

EDIT: hah, i saw that incarnate you post ninja!
Nov 20, 2013 incarnate link
Perhaps make the ships actually correspond to their noted length?

I'm going to assume you mean make the numbers correspond to the ship's actual length, and not make the ships match the numbers.. which would change the flight dynamics of every ship in the game for the sake of a relatively silly stat (and is what I responded to in my initial reply).

Anyway, it's possible to have the client calculate the actual length of the ship and display that data. But it would require a little doing, as there would be C++ code that needs to be modified to do that, as Lua is kept separated from low-level functionality like this (for performance and memory usage reasons). It has never been very high on my priority list, and honestly.. it still isn't. Someone complains that the length stat is inaccurate about every two years or so, and I really don't care?

But, I do think there is merit to allowing accurate comparison of maneuverability, as TRS asks, so such a calculated-length stat would inherently fall out of creating that kind of feature.

Yeah its funny that all this time we actually thought that the length stat had gameplay impact i'm surprised nobody has asked the question as to whether all the stats actually mean something.

No, people just forget. We've told people the stat isn't real at least half a dozen times. Off the top of my head, it's the only one that isn't real.

Anyway the important thing with the spin ratio is not that it's accurate. Nobody is going to care to calculate how many seconds it would take to turn 90 degrees in a skyprom or whatever, the point is for comparative purposes all we need is a formula that is uniformly applied to all the ships so we can see relatively how fast each turns.

I keep trying to answer that. You will not get such a formula that is able to accurately reflect what you want, without including the distribution of mass in the object. I chose the word "inaccurate" to describe it, perhaps it would be simpler to say: wrong, and potentially mis-informing.

Here's a question, if you expand the mesh with this new dynamic addon feature, you know the one you use for the type-S and stuff, does that mean that hypothetically you could completely change the spin mechanics of a ship by say adding more length dynamically or sticking on a dingbat? If you do this to all variants I spose we're going to get a huge deviation of manuverability amongst variants.

What? We aren't doing any dynamic geometry for the Type S. We're doing dynamic texture compositing.. that has nothing to do with object geometry.

That was the original plan for the dynamic-geometry ships, way back in the day. But we've never engineered everything required to do that. Had we implemented it, though, yes, it would have altered the distribution of mass dynamically based on the ship's specific configuration.

For instance, adding a lot of cargo to a Marauder would have expanded the number of crates behind it, which definitely would have changed the flight dynamics. Inverse kinematic (flexible) links between the crates would have made things even more interesting.
Nov 20, 2013 TheRedSpy link
I keep trying to answer that. You will not get such a formula that is able to accurately reflect what you want, without including the distribution of mass in the object. I chose the word "inaccurate" to describe it, perhaps it would be simpler to say: wrong, and potentially mis-informing.

Yeah don't worry I understood what you meant and I agree it isn't that important.

How this could be done is by creating a program that moves the mouse on the screen or sets the position of the ship 90 degrees and records how quickly it takes for the auto-aim recticle to reach that spot on both the x and y axis, then you could plot two series on a graph for x rotation and y rotation and compare like so. That factors in the mesh of the ship just fine, the devil would be in the detail of how exactly you record it.

What? We aren't doing any dynamic geometry for the Type S. We're doing dynamic texture compositing.. that has nothing to do with object geometry.

But the images of the Type S show that it is both longer in the front and has two extra engine pylons on the side.. that has no effect on its spin maneuverability whatsoever?
Nov 20, 2013 Pizzasgood link
You're just confusing terminology. What Inc meant is that they aren't implementing anything right now that would let you buy a ship and then do stuff that alters its model. Yes, the Trident Type S will have a differently shaped model (and therefor different turning), but there isn't anything "dynamic" about that.
Nov 21, 2013 TheRedSpy link
gotcha
Nov 21, 2013 abortretryfail link
The actual length of the ship does have an impact, as it's a factor in the actual distribution of mass throughout the ship (mass is evenly distributed, but within the respective geometry).

Is the Tunguska Centaur Aggresso's collision mesh actually longer than the Centaur MkIII then? As far as in-game comparisons go, they seem to be exactly the same size, but you've already said before that the collision mesh can be bigger than the visual one, like in the case of the old Vultures.
Nov 21, 2013 incarnate link
How this could be done is by creating a program that moves the mouse on the screen or sets the position of the ship 90 degrees and records how quickly it takes for the auto-aim recticle..

Yeah.. that's not what we would do. We have the actual math. But I guess if someone wanted to do it via a plugin or something..? That'd probably be ugly.

Is the Tunguska Centaur Aggresso's collision mesh actually longer than the Centaur MkIII then? As far as in-game comparisons go, they seem to be exactly the same size, but you've already said before that the collision mesh can be bigger than the visual one, like in the case of the old Vultures.

All the Centaurs use the same collision meshes. Pretty much any given ship type is going to have the same collision mesh (and probably the same visual mesh, the "new vulture" being the obvious exception).

Yes, the Vulture's collision mesh is a little tweaked compared to the visual, but by and large.. for the purposes of this discussion, they match up pretty closely. Closely enough that I'm not sure you would really notice the physics/handling impact of making it match up "exactly".